SDG Keyword Lists for Curriculum Classification

SUPPLEMENTARY MATERIAL

Supplementary Keyword Dictionary

SDG Keyword Lists for Curriculum Classification

S1. Overview

This supplementary material presents the keyword dictionary used for the classification of Sustainable Development Goals (SDGs) within undergraduate accounting curricula in Brazilian Higher Education Institutions.

The dictionary was developed to support the automated identification of SDG-related content through computational text analysis applied to course syllabus descriptions (ementas). The keyword lists were incorporated into the Python-based classification procedure described in the main manuscript.

The construction of the keyword dictionary was primarily informed by the Sustainable Development Solutions Network (SDSN) taxonomy, official United Nations SDG terminology, sustainability literature, educational policy documents, and prior studies employing text-mining and content analysis approaches.

This supplementary material aims to ensure methodological transparency, replicability, and scalability of the proposed analytical framework.

S2. Structure of the Keyword Dictionary

The supplementary file contains a structured list of keywords associated with each of the 17 Sustainable Development Goals (SDGs).

For each SDG category, the database includes:

SDG number

SDG title

Associated keywords and thematic expressions

Synonymous and context-related terminology

Sustainability-related concepts identified in the literature

The keyword lists were organized to maximize the identification of sustainability-related content across curricular documents while preserving thematic consistency with the United Nations Sustainable Development Goals framework.

S3. Development of the Keyword Lists

The keyword dictionaries were developed through an extensive review of:

Sustainable Development Solutions Network (SDSN) terminology and thematic dimensions

Official United Nations SDG documentation

Sustainability and accounting literature

Educational policy documents

Prior studies employing text mining and content analysis approaches

The lists incorporate:

Conceptual terms

Policy-related terminology

Sector-specific expressions

Synonyms and thematic variations

Terms associated with sustainability, governance, inclusion, environmental management, economic development, and social responsibility

The inclusion of multiple linguistic and thematic variations aimed to improve the sensitivity and robustness of the classification process.

S4. Classification Procedure

The keyword dictionary was integrated into a Python-based text classification routine used to identify potential alignment between course syllabus descriptions and the Sustainable Development Goals (SDGs).

The procedure involved:

Text preprocessing and normalization

Removal of formatting inconsistencies

Keyword matching across syllabus descriptions

Assignment of SDG categories based on identified terms

Generation of structured classification outputs for subsequent analysis

Because some keywords may be associated with multiple SDGs, the classification process considered thematic proximity and contextual interpretation to reduce overlapping classifications and false positives.

The identification of one or more keywords associated with a specific SDG was considered evidence of potential curricular alignment with the corresponding sustainability objective.

S5. Analytical Use

The keyword dictionary supports:

Automated SDG classification of curricular content

Identification of sustainability-related themes in higher education curricula

Large-scale curriculum analysis

Comparative institutional analysis

Replication of the methodological framework in different educational contexts

Adaptation of the classification model to other academic disciplines and regions

The structure was specifically designed to support scalable and reproducible applications of computational sustainability assessment in education research.

S6. Reproducibility Note

This supplementary keyword dictionary is intended to support methodological transparency and reproducibility.

Researchers may adapt, expand, translate, or refine the keyword lists according to:

Linguistic specificities

Regional contexts

Disciplinary differences

Institutional characteristics

Emerging sustainability terminology

Future applications may also incorporate machine learning, semantic analysis, or natural language processing techniques to complement the keyword-based classification approach proposed in this study.